Spine Surgeries
Background
Spine surgeries often use fixations and implants attached to vertebrae using screws. It is important to place the screws properly so they do not touch or violate neither spinal cord nor arteries. It can be a difficult task due to the needed precision, high density and constrained access to the vertebrae. For these reasons surgeons use support systems that can enhance the accuracy of the screw placement.
In spine surgeries there are the following methods used for placing the screws:
1. Purely manual
2. Manual using navigation systems
3. Using robotic systems
Manual Methods
In the traditional manual technique, a surgeon on the basis of the pre-operative CT scans visually judges the screw trajectory. During drilling, the fluoroscopic images are taken to verify if the trajectory is correct. An advantage of this technique is that except standard reconstruction systems no additional tools are needed and it can be always used in case of an emergency. On the other hand it strongly relies on the surgeon's experience and can be subject to his changing predisposition. Security is also doubtful as the fluoroscopic images are taken only after the drilling is done. The accuracy and information shown on those images can also vary. Drilling is technically difficult because the tools are held in hand. Surgeon needs to have a very good coordination and be able to simultaneously do many tasks. Due to those disadvantages a screw misplacement rate on the level of 30-50% in the cervical spine was reported.
Manual Methods Using Navigation Systems
Navigation systems can measure the position of surgical tools and a patient in the operating room. Currently most often the optical tracking is used for measurements but other methods such as electro-magnetic tracking can be used. Procedures involving those systems will be referred as the image-guided surgeries. Because of the improved accuracy image-guided procedures made the screw placement in the cervical spine possible for certain patients. The image-guided surgeries in the spinal domain are still done manually. For this reason the surgical tools though tracked can be wrongly positioned because of the human constraints. Precision can be a subject of a variable human factor. These techniques demand increased attention from the surgeon as he needs to coordinate operations with virtual indications on the screen. In case of a procedural error big inaccuracies can appear and for this reason a staff training is important. Problems with the verification of the registration accuracy are common
Methods Using Robotic Systems
Few attempts have been done to introduce robotic systems for spinal surgeries. One of them is developed at the German Aerospace Center (DLR) Miro/KineMedic robotic system. It is designed for a surgical telemanipulation. The robotic part of the system consists of three lightweight robotic arms. Each joint is equipped with a force sensor and uses a sophisticated control system with the force feedback and the gravity compensation. The robot's redundancy is used for the workspace optimization and allows to fulfill additional criterias in the operating room. Proposition of the possible setup for a pedicle screw placement with the Miro/KineMedic system would consist of the DLR lightweight robotic arm, an optical tracking system and the software. The surgeon plans the surgery in advance. In the operating room several robot control modes are available. Initially the robotic arm is moved to the planned position by the surgeon using a hands-on impedance control. When it is in place, the surgeon can start drilling using a driller held by a passive tool holder attached to the robot's end effector. The robot compensates for the position errors while surgeon does axial movement. Authors do not specify in which parts of a spine the robot could work. The proposed registration method using a surface matching only could be insufficient in a general situation as those algorithms need a good starting point and converge to the closest local minimum. It is not specified if in this system standard surgical reconstruction tools could be used which can be crucial for the acceptance in the medical domain. A relatively big robotic arm can have disadvantages in a dense environment of an operating room. It is not said how it would be interfaced with the equipment of an operating room. Sophisticated impedance-control algorithms can be difficult to certify in the medical domain and till now no such arm was certified. Expected accuracy of the system is not mentioned. Accordingly to the author's knowledge no further publications concerning this proposition are available.
Other robotic system for the spinal surgery is the Mazor's SmartAssist. It consists of a miniature robot attached to the spine with a base platform and a workstation for planning and navigation. Registration is based on the matching between pre-operative CT scans and intra-operative fluoroscopic images acquired with a calibrated device. In the next step the robot moves to planned spacial position and the surgeon performs a surgery via the tool guide. The robot does not move during the intervention acting as a tool holder (passive guidance). The system was tested with good results. The SpineAssist can be used only in the thoracic and lumbar parts and can not be used in the cervical spine where high accuracy is most important. Fluoroscopic registration has certain disadvantages and needs a calibrated C-Arm. Possible hard to detect errors were reported. The robotic arm does not compensate for random vertebral movements while drilling. Drill slippage on the surface of the vertebrae causing big inaccuracies was reported.
Another robotic system for spinal surgery is the Cooperative Robotic Assistant. It consists of a 6 degree of freedom robot with a kinematically closed structure. It uses a new drill-by-wire mechanism for placing the screws and uses a 1 degree of freedom haptic device to provide the force feedback for the surgeon. Achieved accuracy below 1 [μm] of the robotic part was reported. Authors claim that closed construction was chosen for rigidity reasons. The robot is taking a lot of space in the operating room. Equipment of the operating room should be strongly adapted to be used with this system. The drill-by-wire mechanism needs its own tools which can be a limit for acceptance in the medical field. The system does not perform any external measurements so nothing about registration methods is known. The precision of the registration will strongly influence the accuracy of the robotic arm measured separately. Other robotic system is the Spinebot system for the lumbar spine surgery. It consists of a 3 degree of freedom positioner, gimbals and drilling tool having 2 degree of freedom each. It uses an optical tracking system for registration and measurements. Big advantage of the system is that during the surgery holes in spine can be drilled percutaneusly (through the skin). The system can work only in lumbar part of the spine. In this area needed accuracy is much lower than in cervical part and access is easier.